The mystery of Polls.

While struggling with the statistical evidence for the second paper, I found this article in the Globe and Mail about the good sides of polling. It is interesting as it provides arguments for why polls foster democracy.

As a poli sci student polls have always been a topic on which I had ambiguous and paradoxical feelings.

In my first year I remember my ‘introduction to research’ prof telling us how bad and biased polls are. They were pictured as the new opium of the media, creating public opinions where there were none and transforming the pure will of separate individuals into an aggregated meaningless and manichean reality. This was reinforced by my anthropology course and political sociologists that emphasized the necessity of qualitative research and interviews to catch the complexity of human behavior, that, again, cannot be reduced to numbers and percentages.

Then, came my statistics course in second year. The art of probabilities and their capacity to produce relatively true assesments of reality based on a pretty small sample of the population. However, my teacher took a lot of time to distinguish good statistics from bad statistics and the latter were usually found in newspaper and in the media. Therefore, my faith in polls was growing theoritecally but I was still skeptical about what I was reading in the news.  What I was tought is that apparently some people know how to do polls but those people do research and publish in scientific journals and not in the media or not too often. Too bad.

So, are polls of any use? Are they completely biased, unprofessional and non-orthodoxical statistically? Well, apparently they prove to be not that far from reality when it comes to elections. Pretty mysterious.

Mr. Keeter hold out hope: In the collective sampling of public opinion, we become rational. The averaging of diverse perspectives, he says, offsets the errors of the uninformed – which sounds like a pollster’s definition of true democracy.

This is where my skeptical side jumps in and thinks: the conclusion could as well be that polls reflects a reality that is imperfect and so does the votes. Both are biased in some ways, maybe they are biased in the same way? Is there a way to know the public will? Does such a will even exists?

And that is when I read myself and think that after four years of political science my judgement is biased. As political scientists we try to understand the social reality as well as we show how complex and impossible it is to understand it. Once that acknowledged there is only one thing to do: pick a side. Choosing between being an idealist and a skeptic. If I choose the second option, I answer no to any question and stop trying to understand the world. If I choose the first option, I might be wrong but at least I will have tried to understand something, which cannot be a bad thing after all.

As I chose the  first option I guess I have to believe that we can create useful polls. They might not be perfect but as long as we acknowledge their biases and do not take them as the mirror of the public will, they are a way to measure ‘something’. This something can help us understand ‘something’ else which might take us closer to reality.

(This is the result of a 4 years degree in falsificationism, aka Political Science)

One response to “The mystery of Polls.

Leave a Reply

Your email address will not be published. Required fields are marked *